Stream: implementers
Topic: value or child or Null?
Alexander Henket (Jun 19 2018 at 12:43):
We have V3 input that we create FHIR output for. Sometimes this input has NullFlavors present with an explanation of why the value is not present. If I don't have data, but I know why, I'd like to state that.
In FHIR an element SHALL have a @value or a child element and the .Net validator errors on a construct like this.
<route>
<extension url="http://hl7.org/fhir/StructureDefinition/iso21090-nullFlavor">
<valueCode="NAV"/>
</extension>
</route>
The official invariant in FHIR will not bark, because extension is a child element, but the .Net validator will.
ele-1: All FHIR elements must have a @value or children (expression : hasValue() | (children().count() > id.count()))
Who is right? @Ewout Kramer, @Grahame Grieve?
Lloyd McKenzie (Jun 19 2018 at 12:54):
Including an extension means "route" has a child. And that child has a value. So the invariant is satisfied. If the .Net validator complains, it's in error.
Grahame Grieve (Jun 19 2018 at 13:43):
yes. .net validator is wrong. but I think the problem is deeper than that around an edge case issue in FHIRPath that is clarified in the forth coming FHIRPath spec
Alexander Henket (Jun 19 2018 at 13:59):
@Grahame Grieve does that mean that R4 will raise an error with that change implemented, or will the current status quo remain in place (@value or child, regardless what type)
Grahame Grieve (Jun 19 2018 at 17:46):
umm i'm not sure
Ewout Kramer (Jul 02 2018 at 10:00):
The only reason the .NET validator would draw wrong conclusions here is if the FHIRPath statement is incorrect - at least as far as I can see in the code. Now, there have been some problems with ele-1 since the publication of STU3.
I have heard that we have released some tools with an older version (pre-technical update) of STU3, which had an incorrect FHIRPath statement for ele-1. Also, the current one still is incorrect (it uses a | where it really means or). @Alexander Henket: if you could give me more details about when this exactly occurs, I can probably determine the cause.
Alexander Henket (Jul 02 2018 at 10:12):
This is Simplifier validation. JIRA SD-138. @Arianne van de Wetering Do you have a full example for Ewout
Arianne van de Wetering (Jul 02 2018 at 13:48):
This is Simplifier validation. JIRA SD-138. @Arianne van de Wetering Do you have a full example for Ewout
Done, in a private message to Ewout
:wink:
Last updated: Apr 12 2022 at 19:14 UTC
